Mr Pink57

We've talked about the TJMax settings that Intel has released and they have debunked their own values. I have now just gone by the Distance to TJMax (DTS) vs actually reading the temp.

For instance right now as I type this my computer is at 54c to TJMax, no matter what you change your TJMax to this number never changes.

EDIT: TJMax of 70 puts me at a 16c which is about 60F... My house is warmer then 60F :/ And my proc is at 3ghz with 1.4v.

I strongly suggest you put a disclaimer for yourself on this.

pink

__________________
http://valid.canardpc.com/show_oc.php

crashnburn_819

Awesome info that I had to look up several times, now all in one convenient place! repped

I really only check my Distance to TjMax to be >20 to consider within reason. Is that gonna work or will I be throttling?

ChickenInferno

The actual values for when Core Temperature is approximately equal to TJ Max has never been "debunked." If you read the Other Side of TJ Max Section in my original post you will see that I clearly state how inaccurate idle readings are. So a 16C idle is very inaccurate exactly like I said it would be, and which you are claiming "debunks" my post. TJ Max does nothing for idle but give you inaccurate values and then it only gives accurate values for load when you are close to the TJ Max value.

Since your thermal spec (tcase) is 61.4C and your TJ Max is 80C, then keeping a Dist to TJ Max of 20C would be more than enough to keep perfectly safe. You shouldn't really be throttling at all until you hit around 80C (Dist to TJ Max=0) or not at all if disabled in the bios.

__________________
I edit my posts alot. So please reread my posts if I've edited them; It's for your benefit not mine.

The Truth about Temperatures and Voltages
Flash your Nvidia card with a USB Flash Drive
Start Folding Now

Blameless

That isn't how it works.

The thermal spec (tcase and tjmax are two different measuring points. You should reach thermal spec/tcase the same time you reach tjmax.

And spontaneous, or even overly rapid, damage won't occur at tjmax, that's jsut whent he chip will start to throttle (if thermal management features aren't disabled).

ChickenInferno

The thermal specification (this example E6300 L2) is 61.4C so at 61.4C according to Intel you can start doing damage. If you go beyond thermal specification all the way to TJ Max, then all this is doing is simply going beyond the accuracy point. Temps above TJ Max become more and more inaccurate because it's just not designed for those temps. TJ Max says nothing about damage only about the accuracy of the temps. Since Intel has said that thermal specification is 61.4C, then at any temp at or above 61.4C damage may or may not occur spontaneously or fast. No one can easily predict how heat damage will occur on a cpu unless you do something crazy like heating it to 100C with a max temp or 71C (Obviously it would be fried if left for any considerable amount of time)

I did however make a booboo by saying the throttling temp was 61.4. I didn't reread it.

Alot of information was learned after this post was written. Please refer to the first post for the correct information.

__________________
I edit my posts alot. So please reread my posts if I've edited them; It's for your benefit not mine.

The Truth about Temperatures and Voltages
Flash your Nvidia card with a USB Flash Drive
Start Folding Now

hogans

This has been debated several times, what I have found is this:

Testing my QX6850 at 3.6 GHz. with 1.352v using Orthos small FFT's will give these readings:

Ambient Room temp. = 21c.

Core Temp. using the 80c. TJ Max setting = 39c. Full load.

Core Temp. using the 100c TJ Max setting = 59c. Full load.

My conclusion is, 39c. full load is highly unlikely, 59c. full load seems more accurate to me.

__________________
http://www.overclock.net/intel-mothe...48-thread.html

kpo6969

E8400 CO @ 3.6ghz (TjMax 100c)
Calibration for RealTemp (TjMax was 95c when I did it)



Sensor Test w/ RealTemp (100c TjMax)



CoreTemp, RealTemp, and HWMonitor Temps



My only question in your thread is where do you get your calibration figures from, how, and did you need to adjust?

My example no calibrations needed, I have good sensors (but a lousey o'cr)

ChickenInferno

Okay, looking at the error chart as an example. Your first reported temp is 39C which since it is less than 50C, according to Intel you can only think of this as "a temp lower than 50." Most of my post is not about lowering your temps by changing the TJ Max to something that artifically makes it look like you are running cooler, but that until you are relatively near your TJ Max your values are just very inaccurate. Just looking at your distance to TJ Max value, at a distance of 41 any reading is going to be highly innaccurate regardless of what temperature it is reporting. This could be 39C plus or minus 20. By artifically calling the TJ Max 100C, you raise your reported Temp but the distance is still the same so the innaccuracy is still the same.

The values for TJ Max were found actually on OCN posted by your link to Tom's Hardware when you posted in this thread.

I only had to adjust Core Temp, because Real Temp and HWMonitor were already set to TJMax of 90C.

__________________
I edit my posts alot. So please reread my posts if I've edited them; It's for your benefit not mine.

The Truth about Temperatures and Voltages
Flash your Nvidia card with a USB Flash Drive
Start Folding Now

ChickenInferno

Edit: Eww Double Post sorry

__________________
I edit my posts alot. So please reread my posts if I've edited them; It's for your benefit not mine.

The Truth about Temperatures and Voltages
Flash your Nvidia card with a USB Flash Drive
Start Folding Now

kpo6969

So you never actually calibrated your cpu, you just made all the TjMax the same.

Calibrating your cpu for RealTemp:

http://www.techpowerup.com/realtemp/docs.php

Calibration
The only way you are going to get some reasonably accurate reported temperatues is if you take the time to check and adjust your calibration. On the XtremeSystems forum, rge did some extensive testing with both an IR thermometer as well as mounting a calibrated thermocouple into the IHS lid over top of his processor. The calibration procedure involves running your CPU at a fixed MHz and at a reduced core voltage. This helps equalize the heat ouput of a wide range of CPUs so the difference in reported temperature between a Q6600 G0, an E8400 C0 and an E2160 with 1MB of cache should only be +/- 1C at these settings with the E8400 representing the mid point.

You need to set your CPU to default MHz for this test. The front side bus should be set to 266 MHz for 65 nm processors and to 333 MHz for the newer 45 nm processors. Core voltage needs to be set to 1.10 volts as reported by CPU-Z. The CPU multiplier needs to be set to 6.0 at idle or 12.0 for Core i7. You can either enter these values manually into your bios or you can also enable EIST, Enhanced Intel Speed Step, which is designed to drop your CPU close to these values automatically at idle. Not all motherboards properly support EIST so it's best to check using CPU-Z to confirm these values.

Open your computer case and compare your reported idle temperatures to your room temperature near your case or to your water temperature if you are water cooled. Based on the type of CPU cooling you are using, Core 2 based CPUs should be seeing reported temperatures similar to what rge saw during his testing.

COOLING...............................IDLE DEGREES ABOVE AMBIENT
High end water...............................6C above ambient
High end air (true push/pull) ........6-7C
High end air (1fan).........................7C
Mid air (zalman 9500)....................8-9C depending on fan rpm
Intel stock cooler...........................10-11C

If your reported temperature is too low then you will need to go into the RealTemp Settings window and use a positive calibration factor to increase your reported temperature. If your CPU is reading too high, you will need to use a negative calibration factor. On a multi-core processor during this test, you should have equal temperatures for all of your cores. Though this calibration test is done when your processor is idle, the calibration settings entered will improve the accuracy of your reported tempertures from idle to TJMax. Return your computer to your normal MHz and core voltage settings and if you need to, you can make some further minor adjustments to equalize your core temperatures at idle. You shouldn't need to make any major adjustments at this point.

Based on Intel's specs and during initial testing, the Core i7 CPUs seem to be more efficient and create less heat at idle compared to previous Core 2 Quad processors. Your results will likely be a couple of degrees less than the numbers listed in this chart depending on how your bios and Windows power management options are set.

Settings Window
Idle Calibration
Any value between -19.9 and 9.9 can be used to improve the accuracy of your reported temperatures. Refer to the Calibration paragraph above to determine what settings are appropriate for your processor. A unique value for each core can be set. The temperature sensors on each CPU are unique so for best results, RealTemp should be calibrated to your CPU.